Том 27, № 2 (2017)

An ensemble clustering method is proposed that is based on a weight averaged co-association matrix. The ensemble includes various cluster analysis algorithms whose weights are calculated with the use of cluster validity indices. The properties of the ensemble are analyzed, a probabilistic model is described by which the relations between the characteristics of the ensemble and a quality estimate of a decision are determined, and a method is proposed for determining the optimal weights. The efficiency of the method is analyzed by statistical simulation.

This paper proposes a method for approximating experimental data points by the curves representing the solutions of linear difference equations with constant coefficients, in particular, by the curves of the expcos class. An algorithm for finding the coefficients and initial conditions of this approximation is described. The proposed approach minimizes the root mean square (RMS) deviation. The method is tested on some model examples, including the refinement of the beginning of QRS complexes on a three-dimensional ECG loop (in the form of Frank leads).

Accurate urban areas information is important for a variety of applications, especially city planning and natural disaster prediction and management. In recent years, extraction of urban structures from remotely sensed images has been extensively explored. The key advantages of this imaging modality are reduction of surveying expense and time. It also elevates restrictions on ground surveys. Thus far, much research typically extracts these structures from very high resolution satellite imagery, which are unfortunately of relatively poor spectral resolution, resulting in good precision yet moderate accuracy. Therefore, this paper investigates extraction of buildings from middle and high resolution satellite images by using spectral indices (Normalized Difference Building Index: NDBI, Normalized Difference Vegetation Index: NDVI, Soil Adjustment Vegetation Index: SAVI, Modified Normalized Difference Index: MNDWI, and Global Environment Monitoring Index: GEMI) by means of various Machine Learning methods (Artificial Neural Network: ANN, K-Nearest Neighbor: KNN, and Support Vector Machine: SVM) and Data Fusion (i.e., Majority Voting). Herein empirical results suggested that suitable learning methods for urban areas extraction are in preferring order Data Fusion, SVM, KNN, and ANN. Their accuracies were 85.46, 84.86, 84.66, and 84.91%, respectively.

Pre-processing approaches have been widely used in face recognition to enhance images. However, a notably limited amount of research has examined the use of post-processing methods. In this paper, we propose a novel post-processing framework to improve dimensionality reduction methods for robust face recognition. The proposed method does not work on the features directly; it decomposes each feature into different components using multidimensional ensemble empirical mode decomposition and later maximizes the dependency and the dispersion among classes using a Gaussian function. The performance of the proposed algorithm is demonstrated through experiments by applying several dimensionality reduction techniques on two public databases.

The ease in digital imaging has led to a decrease in image fidelity where illegal reproduction of multimedia information has become difficult to detect. The most challenging problem is to protect image copyright against illegal copies. Therefore a watermark detection process is required to verify the owner of the image. This paper proposes a blind algorithm to extract the watermark. This algorithm is robust against noise and geometric attacks for grayscale images. Robust feature points are detected using the Harris Detector. In the embedding stage, sequences are placed into regions located around feature points. The sequences used are PN, Gold and decimal. In the extraction process, image features are re-allocated using the same detector. Each feature point is used as a center of an N × N region. This region is moved horizontally and vertically within its neighboring pixels. Each move should be registered in a matrix as a correlation value of this region with the initial sequence. This procedure is repeated for all feature points till we find all the watermarked regions. The maximum correlation obtained will determine the center of the watermarked region. The proposed algorithm is robust against a wide variety of tests and is compared to other schemes.

A novel integrated technique is proposed for hierarchical vectorization of electrical drawings in document images. Its first step includes recognition of different electrical symbols and their interconnections based on morphological operations and geometric analysis in three well-distinguished subspaces. This is followed by a hierarchical analysis for detecting the (series-or parallel-connected) super-components in an iterative manner. Finally a compact collection of circuit adjacency lists is produced, which are reduced further by binary encoding. Reconstruction algorithm has also been explained to merit the overall efficacy of the vectorization. Experimental results have been furnished to demonstrate its efficiency and robustness.

Image enhancement by preserving the original brightness is the main challenge of the consumer electronics field. This paper concentrates on the modification of traditional histogram equalization (HE) method to increase its brightness preserving ability. In literature mainly two types of variants have been proposed based on segmentation and modification of the histogram. Both variants are able to preserve the original brightness to some extent. Actually the efficiency of these variants of HE depend on the optimal segmentation and proper modification of histogram. This study concentrates to prove that histogram segmentation or modification which one is better to preserve the image’s original brightness and how much it can be preserved by using the combination of two variants. New hybrid variants of HE method have been proposed in this paper to prove that fact. Results of the proposed methods have been analyzed visually and mathematically.

In this paper, we propose a novel Retinex image enhancement approach by adopting two important prior named brightness channel prior (BCP) and change of detail (CoD) prior. We first derive a rough illumination map estimation method via BCP and Retinex model. Then, we present a combination refining method which involves the guided filter and a new total variation (TV) smoothing operator to eliminate the block effect in the rough illumination map while maintain the local smoothness property. In addition, we propose a novel sharping algorithm rely on CoD prior to improve the visual effect of the degraded image. Experimental results verify that our approach outperforms current approaches in terms of effectiveness, efficiency and universality.

This paper starts a brief historical overview of occurrence and development of fuzzy systems and their applications. Integration methods are proposed to construct a fuzzy system using other AI methods, achieving synergy effect. Accuracy and interpretability are selected as main properties of rule-based fuzzy systems. The tradeoff between interpretability and accuracy is considered to be the actual problem. The purpose of this paper is the in-depth study of the methods and tools to achieve a tradeoff for accuracy and interpretability in rule-based fuzzy systems and to describe our interpretability indexes. A comparison of the existing ways of interpretability estimation has been made We also propose the new way to construct heuristic interpretability indexes as a quantitative measure of interpretability. In the main part of this paper we describe previously used approaches, the current state and original authors’ methods for achieving tradeoff between accuracy and complexity.